Abstract: The construction industry is one of the most important socio-economic sectors of the Spanish economy and one of the most affected by workplace accidents. An analysis of the data on accident rates is needed, in order to identify variables related with workplace accidents and to define the measures that need to be taken for their reduction. In this study, an analysis is conducted using Bayesian Networks and data from the 7th National Survey on Working Conditions (VII NSWC), to study the relations between workplace accidents, visiting a doctor for occupational reasons, time in the company/sector, information that workers have on workplace risks in the workplace, and information and training on workplace risks that workers have received over the past two years. The NSWC survey, which is conducted every four years, was administered to 8892 workers, in Spain, in 2011. The values derived from the analysis yield certain implications involving the aforementioned variables and how to reduce the probability of workplace accidents. From among the variables under study, information on workplace risks is the most important, with the probability of suffering an accident in the construction industry doubling when such information is insufficient. In accordance with the results, these implications could also help with decision-making focused on improvements to training and on-the-job information, intended both to prevent and to reduce workplace accidents.